From the potential of orbital railguns, to space elevators on the moon and Mars, to the threat of AI taking over your job, to the latest on Neuralink, today’s lightning round video features questions from Patreon supporters. Thanks for the great questions guys!
Could we get to space with a railgun?
Could your job be replaced by AI?
What’s up with Neuralink?
Could we build a space elevator on the Moon?
Why aren’t plug-in hybrid cars more popular?
All this and more in today’s Lightning Round video.
Lightning round video, by the way, not to be confused with round lightning video. For that you need to see my video on ball lightning.
But let’s get into today’s questions…
The idea of EM-driven mass accelerators as an orbital launch platform has been around for some time, yet seems to rarely get past the drawing board. Has there been much progress in designing such a platform? What major issues might still need to be overcome?
Weirdly, this came up in the OLF reunion livestream we did a few weeks back, you can check that out if you want to hear me talk about it with Tim Dodd and Ben Sullins.
Probably the biggest issue is just the size of the launcher because it has to get out of the atmosphere.
Rockets only go straight up for the first 30 seconds or so, just high enough to get past the thickest part of the atmosphere and then they have to reach orbital velocity which is 15,000 miles per hour.
You know how things in orbit burn up in our atmosphere when they come back down? That’s what would happen on the launch pad if you were to launch from the ground at that speed.
So an orbital railgun would have to be at least 150 miles long and get up to several miles into the air. This is megaproject territory.
Some have suggested putting the launch ramp up the side of a mountain but it would have to be up a very high mountain to get into thin enough atmosphere and that would also make getting the payload to the launch site more difficult.
Not to mention the entire length of the railgun would need to be in a vacuum tube, which would take a lot of energy…
You’re basically building a vertical hyperloop.
Other option is rocket assisted railgun where it yeets at lower velocity so it doesn’t vaporize right off the launchpad and then once it reaches altitude a rocket takes you the rest of the way and reaches orbital velocity.
Which is an interesting idea, it’s kind-of like the railgun takes the place of a first stage of the rocket.
Now I couldn’t find any companies that are currently working on this kind of thing, feel free to point out any I missed in the comments, but there is one worth working on a similar idea… many of you probably know where this is going…
Spin Launch systems recently tested their prototype megayeet machine that spins a payload at high speed before launching it out of a tube, letting it get to extremely fast speeds without the need for a long railgun track.
They call it an orbital accelerator – a kinetic space launch system
Totally electric and reaches hypersonic speeds but still would need a rocket assist because again, orbital speed would melt everything.
In fact as Scott Manley pointed out in his video on this, thermal imaging on the projectile shows how hot it gets immediately after launch from air friction.
There’s still a big question mark whether they’ll be able to reach orbit with this, but I guess we’ll see how things go.
Back to railguns, they did have a bit of a setback recently because the US Navy ceased development of their hypersonic railgun after spending $500 million dollars on it.
They had a prototype on board the USS Trenton but they’ve dropped the program in favor of a hypervelocity projectile, or HVP, which can be fitted to traditional ship guns.
So, I’m skeptical that we’re ever going to see a real orbital launch railgun, though I have said that there’s some opportunity in the future on the moon because you don’t have that pesky atmosphere to contend with.
On that note…
Cole Parker – October
Would it be much easier to build a space elevator on Mars or the moon due to the much lower gravity?
So I covered space elevators a while back on my channel – link below – and this was one of the assertions that I made in that video, that it might make more sense on the moon. Because of the lower gravity. I feel like I just said that with the last story. That’s weird…
But yeah, in a nutshell for anybody who doesn’t know, a space elevator is a way to cut down on launch costs because you literally would build a tether from a base on Earth to another base out in geostationary orbit that stays in place through centripetal force.
This is obviously a ridiculously long tether that would need to withstand some major forces, requiring a materials technology that doesn’t exist yet.
Not to mention it would need to survive getting pelted by micrometeoroids and space debris. And satellites.
It’s a tough sell.
But, the advantages if we could do it would be huge, getting a payload into orbit would be as simple as taking the elevator. Pretty huge deal.
Many of these major challenges would not exist on the moon. And it turns out there has been a lot of thought put into this.
In a 2019 paper, two astronomers from Cambridge and Columbia Universities took a detailed look at this problem and did the math. And the math… checks out.
It wouldn’t be quite like a space elevator on Earth because the moon doesn’t spin like the Earth does, being tidally locked and all.
Instead, the idea here is that you could build a tether from the surface of the moon all the way down to basically geostationary orbit above Earth, where you would have a docking station. This would be pulled by Earth’s gravity to keep the tether taut.
But yeah, to get to the moon, you would just fly up to geostationary orbit and dock at the station, then just ride up to the moon powered by solar energy. It would lower the costs by a factor of thousands.
And, according to the authors, might only cost a few billion dollars to build.
Of course, it’s all easier said than done, we’re talking about a 225,000 mile tether. You would have to build a tether long enough to wrap around Earth 28 times. And then launch that to the moon. And then launch it to Earth
Actually, you would probably set up a base in the Earth/Moon Lagrange point and then extend the tether out in both directions from there. But still.
You would also have to contend with the fact that the moon’s orbit is slightly elliptical so that counterweight hanging over Earth would rise and fall throughout the month.
So yeah, the gravity helps, but the insane length of the tether kinda raises the difficulty level back up. Still, there are companies working on this like Liftport.
But you also mentioned Mars, and that’s where things get interesting because yeah, with 1/3rd the gravity and 1% the atmosphere, there’s a lot less stress on the line, but Mars has another ace up its sleeve. It s moon Deimos.
Deimos has a rotational period shockingly close to that of Mars’ sidereal day (30.3 hours vs. 24.6 hours) – which is shockingly close to our day – and it’s about the same distance as our geostationary orbit on Earth. (14,576 miles)
And if that wasn’t enough, it’s made mostly of carbon. So if we could establish a carbon nanotube factory on Deimos, we could mine its carbon, build a tether and drop it down to the surface.
We would need to adjust the orbit a little bit though.
Anyway, that presents some interesting opportunities but would require a lot of technologies we still don’t have yet. Interesting to think about though.
Mike Reed – November
Which jobs are the most likely to be replaced as AI grows? How will low-wage manual labor workers get a job once AI does all of this work?
This is a very good question. And I wanted to give it a very good answer. So I found an article on Medium from a guy named Kai Fu Lee, who is an AI Expert, President of Google China, and author of the book AI Superpowers.
In this article he lists the 10 jobs most at risk from AI in the coming years. I mean ultimately all jobs will be at risk, but these are the ones likely to be most affected first.
Number 1 is Telemarketers/Telesales
I mean, we’re already seeing robocalls taking over this space but he talks about how AI can be trained to understand people’s tone and adjust its own tone accordingly, and to even know if you react better to male or female voices and choose the best one for every call.
Similar to that is Customer Support.
Again, we’re already seeing chatbots and AI assistants taking over a lot of this online, but it’s closing in on phone support as well, because a lot of the people doing those jobs are just reading scripts anyway.
Next up is Warehouse Workers.
Not a surprise to anyone, I’m sure, we’re already seeing robots doing a lot of the heavy lifting in Amazon’s warehouses and whatnot, but as they become more dextrous, they’ll continue taking over spots that are now held by people.
Then you’ve got Clerks and Operational Staff
There’s a lot of large companies that deal with huge amounts of data, and they employ thousands of people to manage and process this data. Everything from filing, procurement, inventory management, error correction, estimating sales and reporting. And much of that can already be automated.
Telephone operators are on the chopping block.
With the progress of speech recognition and dialog-oriented speech synthesis, the days of an actual person to answer and direct calls are pretty much numbered for a lot of businesses.
Tellers and Cashiers are going to be a thing of the past.
Think about how many self-service check-out lines there are now. That trend is not going to slow down.
According to Kai Fu Lee, Fast Food Worker jobs are in trouble.
Basically any repetitive and stationary job is likely to be done by a robot soon and burger flipping is pretty much the definition of that. There are already robotic burger flippers in some restaurants. Also order taking is something that can be automated as well.
Sticking with restaurants, dishwashers will be going away soon.
There is already a startup in California called Dishcraft developing systems that can wash thousands of dishes a day, using far less water and energy.
Something else likely to be phased out are Assembly Line Inspectors.
Again, it’s something fixed and repetitive. And computer vision can easily find cracks and blemishes in units on an assembly line, probably better than a human.
And last but not least are couriers.
Delivery robots and drones are likely to take over this space, especially in structured environments like large office complexes and hotels.
Basically if your job is repetitive and requires little human interaction, you might want to start looking for something that’s more human-focused, requires more dexterity.
If you’re in one of those jobs and are wetting your pants at this very moment, I’ll link to his article down below, he does give some tips for ways to pivot into something that’s more safe in the long run. https://kaifulee.medium.com/top-10-most-endangered-jobs-from-ai-4675e0eff1a8
Brian Beswick – October
What’s up with NeuraLink???
Yeah, I’ve been asked to follow up on that topic for a while and honestly there just hasn’t been that much to report on? But here’s some news from this last year or so.
Last time we got a presentation, Elon showed how the implant was working in a pig’s brain to pick up on signals from its snout.
The pig wasn’t controlling anything, it was basically just showing how they were receiving signals from the snout whenever it touched something.
There was also an updated version of the implant that looked a little rougher and clunkier than the first presentation, but at least it was a real thing and not just a rendering.
in April of this year, they revealed footage of a monkey playing pong using a chip in its brain.
They show how they trained the monkey, whose name is Pager, to play this little computer game where he boops the squares using the joystick, and the neuralink chip is registering the signals being created to do this. It even streams to a smartphone.
By the way, that metal tube in his mouth is his little reward mechanism that gives him some food every time he gets it right.
But they eventually collect enough data that they can control the game with the implant, as you can see, the joystick is unplugged here, so he’s moving that cursor with his mind.
And then they showed him playing pong, and there’s no joystick at all now, this is totally the monkey and a neuralink chip.
This is actually a pretty big deal because one of the first things they want to do with this chip is make it so that quadriplegics can operate phones and computers with their minds. This looks like they’re pretty close.
So their next presentation, which could be soon, it’s been over a year since their last one – it could possibly have a person using one of these. Place your bets in the comments.
Real talk though, people being able to control basic computer stuff with their brains is nothing new, there have been things like the Utah array around for a while but this would be much smaller and probably more sensitive.
But yeah right after the monkey video, it looked like they might have hit a rough patch because Max Hodak, who was the president and co-founder along with Elon, he left the company. Nobody really knew what to make of that.
But still, they did secure a third round of funding in late July worth $205 million from people like Google Ventures and Peter Thiel.
This makes the total investment into it $323 million, which considering Elon is worth like $300 billion right now… Seems kinda low?
So that’s the last news I could find about Neuralink but there are a couple of other companies doing stuff in this space.
Including a company called Science Corp – most ambiguous name ever – they raised $47 million in August.
Interesting thing about Science Corp, their CEO is Max Hodak… The guy who left Neuralink.
Yeah, apparently this company wants to do brain machine interfaces but they hired a bunch of CRISPR specialists. So there may be some bioengineering or gene manipulation involved.
But they are very much in stealth mode, don’t know much about what they’re working on.
And another company called BrainGate made some news in April this year for wirelessly connecting participants’ brains to a computer.
Basically what Neuralink did with a monkey, these guys did with two people.
So progress is being made in a lot of different directions.
Cole Parker – November
Why aren’t Plug-in Hybrid Electric Vehicles (PHEVs) more popular in America and Europe?
I don’t know that I have a good answer for this, so I’ll just speculate. If someone has a better answer, please share in the comments.
My guess is that EVs are more popular because we now associate electric cars with instant torque and being fast off the line and having cool tech and plug in hybrids are seen more like gas sippers.
And if you want a car that burns less gas… well, you can’t get less than no gas at all.
So EVs appeal to the torqueheads and the tree huggers and PHEVs appeal to…
Not saying that plug in hybrids don’t have their place, they kinda have the best of both worlds, you can commute daily without using any gas at all but still have that gas range for long-distance drives.
But I guess EVs are more popular right now because they’re the new hotness. Which, even 5 years ago was a sentence I couldn’t imagine saying.
And I think a lot of the PHEV models from the last several years were kinda compliance cars, meant to get the fleet efficiency down for regulatory reasons, so car companies haven’t really marketed them very well because they weren’t really trying to sell them… I may be making an assumption there, you tell me.
But yeah, like I just said, the fact that this is even a question is really mind-blowing to me. It’s like, as an EV advocate, it really felt like I was banging my head against a wall for so long… And now suddenly everybody’s scrambling for them, and car companies are promoting them to the point that we find ourselves wondering what happened to hybrids. It’s amazing.
So that’s all the questions we have today, feel free to chime in with your thoughts in the comments and if there are any of these topics you want to see a full deep dive video on, let me know. It could be a thing.